This is a proposal to study formal thought disorder in mid- and late-childhood psychoses. The proposed research portion project will incorporate cognitive and linguistic elements into the Childhood Psychoses Formal Thought Disorder Rating Scale (CPFTD-RS), developed by the PI for an evaluation in children of the DSM-III's formal thought disorder symptoms. Age-related changes in the measurement of formal thought disorder will be assessed in normal and psychotic children, aged 5-12 years; as well as an assessment of the relationship between formal thought disorder and language and cognitive deficits in psychotic children. The occurrence of formal thought disorder will be determined in three subgroups of childhood psychoses: Schizophrenia, schizotypal personality disorder, and childhood onset pervasive developmental disorder. Each of these groups will be equally represented in a sample of 108 boys, and will be compared to 72 normal boys on the CPFTD-RS, the Curtiss-Yamada Comprehensive Language Evaluation (CYCLE), and on cognitive measures of memory, metamemory, attention, metaattention, conservation, and IQ. The purpose of the supervised research and development portion of the proposal is to train the PI in those skills necessary to rigorously conduct the research project. Although the PI has clinical and research expertise in child psychiatry, she needs supplemental training in language development, cognitive development, and psychometrics. This training will contribute to the inclusion of developmental considerations into the CPFTD-RS. The value of the psychometric training is that the methodological and statistical basis of the CPFTD-RS can then be perfected. This proposed research is particularly important for future research on childhood psychoses. It will provide a reliable method for the evaluation of formal thought disorder in children, and will allow the discipline of child psychiatry to overcome a major obstacle to reliable diagnosis of the childhood psychoses.